The study introduces a machine learning-driven system that conducts real-time and batch searches for adverse media. This is essential for identifying non-financial risks that have become increasingly significant for global market institutions.

Noteworthy Points:

• The system provides high-precision adverse news filtering, which is important for KYC and ESG compliance.
• Adverse sentiment analysis and risk encoding techniques are utilized.
• The proposed application is scalable, addressing the needs of various sized institutions.

This work is significant as it offers an automated approach to a traditionally manual and resource-intensive process. It opens doors for further research on the integration of such systems within larger regulatory tech (RegTech) solutions.